Effects of abandoned coal-mine drainage on streamflow and water quality in the Shamokin Creek Basin, Northumberland and Columbia Counties, Pennsylvania, 1999-2001

This report assesses the contaminant loading, effects to receiving streams, and possible remedial alternatives for abandoned mine drainage (AMD) within the upper Shamokin Creek Basin in east-central Pennsylvania. The upper Shamokin Creek Basin encompasses an area of 54 square miles (140 square kilometers) within the Western Middle Anthracite Field, including and upstream of the city of Shamokin. Elevated concentrations of acidity, metals, and sulfate in the AMD from flooded underground anthracite coal mines and (or) unreclaimed culm (waste rock) piles degrade the aquatic ecosystem and water quality of Shamokin Creek to its mouth and along many of its tributaries within the upper basin. Despite dilution by unpolluted streams that more than doubles the streamflow of Shamokin Creek in the lower basin, AMD contamination and ecological impairment persist to its mouth on the Susquehanna River at Sunbury, 20 miles (32 kilometers) downstream from the mined area.

Aquatic ecological surveys were conducted by the U.S. Geological Survey (USGS) in cooperation with Bucknell University (BU) and the Northumberland County Conservation District (NCCD) at six stream sites in October 1999 and repeated in 2000 and 2001 on Shamokin Creek below Shamokin and at Sunbury. In 1999, fish were absent from Quaker Run and Shamokin Creek upstream of its confluence with Carbon Run; however, creek chub (Semotilus atromaculatus) were present within three sampled reaches of Carbon Run. During 1999, 2000, and 2001, six or more species of fish were identified in Shamokin Creek below Shamokin and at Sunbury despite ph as low as 4.2 at Sunbury and elevated concentrations of dissolved iron and iron-encrusted streambeds at these sites.

Data on the flow rate and chemistry for 46 AMD sources and 22 stream sites throughout the upper basin plus 1 stream site at Sunbury were collected by the USGS with assistance from BU and the Shamokin Creek Restoration Alliance (SCRA) during low base-flow conditions in August 1999 and high baseflow conditions in March 2000. The water-quality data were used to determine priority ranks of the AMD sources on the basis of loadings of iron, manganese, and aluminum and to identify possible remedial alternatives, including passive-treatment options, for consideration by water-resource managers. The ranking sequence for the top AMD sources based on the high base-flow data generally matched that based on the low base-flow data. The contaminant loadings generally increased with flow, and 10 previously identified intermittent AMD sources were not discharging during the low base-flow sampling period. The top 3 AMD sources (SR19, SR12, and SR49) on the basis of dissolved metals loading in March 2000 accounted for more than 50 percent of the metals loading to Shamokin Creek, whereas the top 15 AMD sources accounted for more than 98 percent of the metals loading. When sampled in March 2000, these AMD sources had flow rates ranging from 0.7 to 19 cubic feet per second (1,138 to 32,285 liters per minute) and pH from 3.5 to 6.4 standard units. Only 1 of the top 15 AMD sources (SR21) was net alkaline (alkalinity > acidity); the others were net acidic and will require additional alkalinity to facilitate metals removal and maintain near-neutral pH. For the top 15 AMD sources, dissolved iron was the principal source of acidity and metals loading; concentrations of iron ranged from 3.7 to 57 milligrams per liter. Dissolved manganese ranged from 1.8 to 7.1 milligrams per liter. Dissolved aluminum exceeded 3.8 milligrams per liter at six of the sites but was less than 0.2 milligram per liter at six others.

Alkalinity can be acquired by the dissolution of limestone and (or) bacterial sulfate reduction within various passive-treatment systems including anoxic or oxic limestone drains, limestone- lined channels, or compost wetlands. Subsequently, the gradual oxidation and consequent precipitation of iron and manganese can be accommodated within settling ponds or aerobic wetlands. Assuming an iron removal rate of 180 pounds per acre per day (20 grams per square meter per day), constructed treatment wetlands at the top 15 AMD sites would require a minimum area ranging from 0.1 to 17.8 acres (405 to 71,670 square meters). Implementation of passive treatment would not be feasible at most of the top 15 and many lower priority AMD sites considering the proximity of many discharges to streams, roads, or railroads, and the limited availability or access to land at the discharge location. The reduction of infiltration and removal of culm waste and (or) the relocation of the discharge to nearby areas could decrease the AMD quantities and facilitate treatment at some of the priority AMD sites.